225Ann Ist super sAnItà 2007 | Vol. 43, no. 3: 225-228

re

seA

rc

h f

ro

m A

nIm

Al t

est

Ing

to

cl

InIc

Al e

xp

er

Ien

ceElectromagnetic immunity of infusion pumps

to GSM mobile phones: a systematic review Giovanni Calcagnini, Federica Censi, Michele Triventi, Eugenio Mattei and Pietro Bartolini

Dipartimento di Tecnologie e Salute, Istituto Superiore di Sanità, Rome, Italy

Summary. Electromagnetic interference with life-sustaining medical care devices has been reported by various groups. Previous studies have demonstrated that volumetric and syringe pumps are sus-ceptible to false alarm buzzing and blocking, when exposed to various electromagnetic sources. The risk of electromagnetic interference depends on several factors such as the phone-emitted power, distance and carrier frequency, phone model and antenna type. The main recommendations and the relevant harmonized standard are also reported and discussed. From the data available in literature emerges that, for distances lower than 1 m there is a non negligible risk of electro-magnetic interferences, although significant differences exists in the reported minimum distances. Interference effects clinically relevant for the patients are rare. No permanent damage to the pumps has been ever reported, although in several cases intervention of personnel is required to resume normal operation.

Key words: electromagnetic interference, infusion pumps, cellular phone, risk reduction behaviour. Riassunto (Immunità elettromagnetica di pompe ad infusione a telefoni cellulari GSM: un’analisi sistema-tica della letteratura). Interferenze elettromagnetiche su dispositivi medici sono state descritte da vari gruppi. Studi pubblicati hanno dimostrato che le pompe a infusione e le pompe siringa sono suscettibili a falsi allarmi e interruzioni dell’erogazione, se esposte a varie sorgenti di campi elettromagnetici. Questo studio è quello di analizzare i dati pubblicati in letteratura relativamente al rischio di interferenze elet-tromagnetiche dovute a telefoni cellulari GSM su pompe ad infusione e pompe siringa. Le principali raccomandazioni per limitare i rischi associati e la normativa internazionale di riferimento sono inoltre riportate e discusse. Dall’analisi della letteratura disponibile emerge un rischio non trascurabile di interfe-renze elettromagnetiche, quando le pompe ad infusione sono esposte a telefoni GSM a distanze inferiori al metro, anche se significative differenze esistono relativamente alle distanze minime osservate. Non esistono dati relativamente a danni permanenti alle pompe, anche se in molti casi dopo interferenza è necessario un intervento manuale sulla pompa per ripristinarne il corretto funzionamento.

Parole chiave: interferenze elettromagnetiche, pompe ad infusione, telefono cellulare, comportamento di riduzione del rischio.

INTRODUCTIONProblems with electromagnetic compatibility

(EMC) of medical devices have been known for some time in hospitals. Research groups, manufac-turers, and governmental and non-governmental agencies have reported incidents related to electro-magnetic interference (EMI) to medical devices. Some of them had life-threatening consequences, others could have had, others can be considered just a nuisance. From 1979 to 1993 the Food and Drug Administration (FDA) received more than one hundred reports related to EMI. These re-ports prompted the need for an increased attention to medical device electromagnetic compatibility by users, manufactures, and standard organizations. There are several motivations behind the increasing researches and efforts in this field: deaths and severe injuries have occurred due to EMI on life-supporting medical de-

vices; the ambient electromagnetic environment con-tinues to intensify (e.g., mobile phones, wireless local area networks, paging system); use of higher carrier frequencies the medical devices have not been tested for; increase in electronic sensors, actuators, and mi-croprocessors based medical devices (e.g., ventilators and infusion pumps); increased number of patients with electrical active implanted devices (pacemaker and cardioverter/defibrillator); widespread of new EM sources such as anti-theft systems and metal detectors, due to the increased need for security in public areas and buildings.

Interestingly, most of the reported incidents before 1993 involved EMI originated from other sources (e.g., electrosurgical units, other medical devices, power line interferences). In the report of Silberberg, 3% of the reports involved mobile phones and 6% hand-held transceivers. It should be observed that

Indirizzo per la corrispondenza (Address for correspondence): Giovanni Calcagnini, Dipartimento di Tecnologie e Salute, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy. E-mail: giovanni.calcagnini@iss.it.

226 Giovanni Calcagnini, Federica Censi, Michele Triventi, et al.

in 1993 the usage of mobile phones was much less prevalent than today. The large number of differ-ent medical devices, the peculiarity of some of them (e.g., implantable vs non-implantable or diagnos-tic vs therapeutic), and the gravity of the potential consequences in case of EMI make difficult to treat this matter in a unique way. The wide number of po-tential sources of interference and their associated mechanism (e.g., conducted vs radiated) make the problem even more complex. These differences are also reflected in the international standard on EMC for medical devices. According to these standards, three groups of devices may be considered: electri-cal active implantable devices (e.g., pacemakers, im-planted defibrillators, nerve stimulators); life-sup-port devices (e.g., ventilators, external defibrillators, electrosurgical units, infusion pumps, monitors); non life-support devices (e.g., ECG, EEG, ultra-sound scanner, MRI, CT-SCAN).

Since the early studies of FDA, various groups have reported problems attributed to EMI from mo-bile phones with medical devices such as ventilators, external defibrillators, wheelchairs, monitors and in-fusion pumps [1-8]. Prompted by these reports, rec-ommendations to restrict the use of mobile phones in critical areas of hospitals have been issued. These recommendations include either the definition of a separation distance or the total banning of mobile phones from intensive care areas and surgical thea-tres, if not from the entire hospital. In view of the lack of evidence reported in other studies [2, 4], the above mentioned restrictions have been criticized [9, 10]. Since most of the reported EMI with medical devices occur only under worst-case conditions (i.e. maximum emitted power and/or very short distanc-es), and because in several cases the clinical conse-quence might be not significant, the debate whether mobile phones pose a real risk is still open [12].

In this review we addressed the EMI problem of infusion pumps. The reason for focusing on this type of devices are various. First, as mentioned be-fore, a large number of parameters are involved in the EMI problems with medical devices., making a unique approach difficult; volumetric pumps and syringe pumps are commonly used in hospitals both in non-critical (wards) and critical areas (e.g., inten-sive care, surgical theatres, first aid departments). Recently, their use at patient’s home has gain popu-larity; in some cases, a malfunction of such devices may pose a significant risk for the patient.

Evidence of cellphone EMI with infusion pumps were observed and documented by [1, 6, 8, 11] while no effects were observed by Turcotte and Witters [2]. According to Klein and Djaiani [13], infusion pumps are particularly prone to EMI. In 2005, Hahm et al., documented the case of an acute Epinephrine poi-soning due to cellular phone interference with an infusion pump [14]. In a previous study our group carried out an experimental investigations on EMI to infusion and syringe pumps exposed to 900 Mhz and 1800 Mhz GSM phones [15]. A systematic re-

view, focused on mobile phones and technologies used in Australia can be found in [16].

METHODSThe major databases were searched (Medline and

Science Citation Index) using the key words “mobile phones”, “cellular phones” and “equipment” or “medi-cal devices”. From a first list of papers, the research was then refined searching for cited authors and papers.

Studies were considered eligible if published in peer-reviewed journal in English and if included testing of infusion and syringe pumps against electromagnetic interference from mobile phones.

In the published studies, several differences in the methodology used to investigate the EMI problem do exist, and it makes difficult a perform a meta-analysis of the published data. In addition, the dif-ferent standards of mobile phones adopted world-wide likely contribute to the heterogeneity of the studies and of the reported effects. Thus, we did not attempt to draw conclusions, but reported the con-clusions of each author.

RESULTSTable 1 summarizes the main findings of our re-

view. We found 6 studies which included GSM mobile phones and infusion pumps among the devices investi-gated. The percentage of devices susceptible to various kind of EMI ranged from 0% to 58%. Some of the studies investigated a very limited number of pumps, thus an underestimation of the rate of susceptibility may be occurred. In all the studies the maximum dis-tances between the mobile phones and the devices were always relatively short (< 0.5 m). There were only two cases of documented changes in the delivery rates [1, 14]. In all the other cases the effect of the EMI consist-ed in buzzing, alarm sound, changes in the displayed information and pump stopping.

DISCUSSION AND CONCLUSIONSSeveral studies have investigated the susceptibility

of medical devices to EMI from mobile phones. As far as infusion pumps are concerned, the percentage of devices susceptible to various kind of EMI ranged from 0% to 58%, indicating real significant differences in the findings of the various groups. Differences ex-ist among the papers, especially regarding the testing protocol, the mobile phones technology, the handset model and the number of pump tested.

Two main international standards are currently applied for evaluating EM compatibility of medi-cal devices [17, 18]. The IEC-EN-60601-1-2:2003 establishes the minimum immunity levels, as well as the methods for conformity assessment. This stand-ard is mainly intended for manufactures and no-tify bodies, as it requires specialized facilities (e.g., anechoic chambers, radio frequency - RF - signal generators, power meters) and trained personnel. In

227gsm ImmunIty of InfusIon pumps

addition, infusion and syringe pumps must comply with the particular harmonized standard EN 60601-2-24:1998 [19], which requires an immunity level of 10 V/m, in the frequency range 80-2500 MHz.

The ANSI C63.18 is a technical guide developed to aid clinical and biomedical engineers in assessing the immunity of medical devices to radiated elec-tromagnetic fields from portable RF transmitters. According to this guideline, medical devices can be tested in the hospital, and RF transmitters can be selected among commercial equipment used in the health facilities. Recently, the ANSI C63.18 has gained diffusion also in papers investigating EMI in hospitals. Indeed, the use of commercial handsets makes it difficult to compare and reproduce results from previous studies: since most of the EMI phe-nomena occur in the near field regions, the antenna patterns of commercial devices play a rule [15].

Since systematic analysis of the clinical relevance of the observed effects have not been carried out in most of the study, it is difficult to obtain the prob-ability of clinically significant EMI. According to Irnich and Tobish [3], there is no realistic danger for drop controlled, infusion and syringe pumps. The most serious effects reported so far included a re-versal in the motor drive of an IVAC 960 exposed to a GSM phone [1], and an acute epinephrine poi-soning probably due to resetting of the pump to the maximum delivery rate (999 mL per hour) [14]. All the other potentially dangerous effects consisted in shutdown of the pump, or in displaying various er-ror modes and alarms, most of the time requiring an external intervention to be restarted.

The recommendations issued to mitigate the EMI risk include either the definition of a separation dis-tance or the total banning of mobile phones from intensive care areas and surgical theatres, if not from the entire hospital. The various proposed rec-ommendations are mostly based on the precaution principle to the risk minimization, rather than on the results of the published study. The “1-m rule” or the “arm’s length rule”, as well as the total ban from intensive care and surgical rooms are the most suggested recommendations. Education and sensi-bilitazion of medical and nurse staff has been also suggested.

If the GSM phone emitted power is reduced, the risk of EMI significantly decreases, as we demon-strated in a previous work [15]. If an adequate base station signal is present, GSM phones are designed to automatically reduce the emitted power to battery saving. Morissey [20] investigated the feasibility of an improved signal coverage as a mean for reducing the emitted power of GSM phones. He compared the power level fluctuations of a GSM phone while walking through a facility with poor and moderate-good coverage. In the presence of an adequate base station signal, the average power barely exceeded 0.01 W. This level corresponds to 0.08 W peak-power. Although we found that a limitation of the mobiles to peak power levels as low as 0.05 W for 900 MHz and 0.0025 W for 1800 MHz is required for the total immunity of the pumps tested, values lower than 0.08 W would significantly reduce the probability of EMI (< 20%). The limitation of the mobile power may be thus obtained increasing the field coverage by install-

Table 1 | Papers investigating the electromagnetic interference (EMI) of GSM phones with infusion pumps

First authorYear

Number of model

tested

Incidence of EMI

Maximum distance (cm) Notes Author’s conclusions

Medical Device Agency1997 [1]

5932

(54%)100

EMI source included: analog and digital phones, 2-ways radios, and LAN

Mobile communication equipment does present a real risk for medical devices

Restrict use of mobile phones in critical areas.

Irnich1999 [3]

66 28 (42%)

n.a.

Results obtained grouping drop-controlled, volume-controlled and syringe pumps

Medical device must be made resistant to mobile phones

1 m minimum distance recommended

Replace of device with more than 50 cm interference distance

Robinson1997 [11]

10

(0%)- Immunity up to 40V/m Suggested safe distance of 1.2 m

Hanada2000 [6]

60

0%-

EMI source: PHS phones (max 80 mW power)

No interference with PHS phones (the power is ten times lower than GSM)

Morissey2002 [8]

92

(22%)25

Infusion, perfusion and feeding pumps

To mitigate EMI, reduce the emitted power by providing good coverage in the hospitals.

Identify most sensitive devices.

Calcagnini2006 [15]

127

(58%)30 (900 MHz)

30 (1800 MHz)

EMI probability as a function of emitted power also calculated

Limit the emitted power of GSM using in-building repeaters

n.a.: information not available.

228 Giovanni Calcagnini, Federica Censi, Michele Triventi, et al.

ing in the hospital in-building repeaters amplifiers or dedicated mini base-stations.

The problem of infusion pump in the domestic en-vironment has so far received poor attention. The fast developing of information technology and telecom-munication infrastructures in the hospitals makes even harder to develop effective guidelines for EMI mitiga-tion. The harmonized international standard on EMC

of medical devices (IEC-EN 60601-1-2:2003) has been recently revised to cover the frequency band up to 2.5 GHz, and it has also increased the minimum immunity requirements for life supporting devices from 3 V/m to 10 V/m.

Submitted on invitation.Accepted on 24 January 2007.

References 1. Electromagnetic compatibility of medical devices with mobile

communications. London: Medical Device Agency; 1997. (MDA Device Bulletin DB9702).

2. Turcotte J, Witters D. A practical technique for assessing electromagnetic interference in the clinical setting: ad hoc testing. Biomed Instrum Technol 1998;32:241-52.

3. Irnich WE, Tobisch R. Mobile phones in hospitals. Biomed Instrum Technol 1999;33:28-34.

4. Hietanen M, Sibakov V, Hallfors S, Von Nandelstadh P. Safe use of mobile phones in hospitals. Health Phys 2000;79:S77-S84.

5. Barbaro V, Bartolini P, Benassi M, Di Nallo AM, Reali L, Valsecchi S. Electromagnetic interference by GSM cellular phones and UHF radios with intensive-care and operating-room ventilators. Biomed Instrum Technol 2002;34:361-9.

6. Hanada E, Antoku Y, Tani S, Kimura M, Hasegawa A, Urano S, Ohe K, Yamaki M, Nose Y. Electromagnetic inter-ference on medical equipment by low-power mobile telecom-munication systems. IEEE Transaction on Electromagnetic Compatibility 2000;42:470-6.

7. Tri JL, Hayes DL, Smith TT, Severson RP. Cellular phone interference with external cardiopulmonary monitoring de-vices. Mayo Clinic Proc 2001;76:11-5.

8. Morrissey JJ, Swicord M. Balzano Q. Characterization of electromagnetic interference of medical devices in the hospi-tal due to cell phones. Health Phys 2002;82:45-51.

9. Aziz O, Sheikh A, Paraskeva P, Darzi A. Use of mobile phones in hospital: time to lift the ban? Lancet 2003;361:788.

10. Myerson SG, Mitchell ARJ. Mobile phones in hospitals: Are not as hazardous as believed and should be allowed at least in non-clinical areas. BMJ 2003;326:460-1.

11. Robinson MP, Flintoft ID, Marvin AC. Interference to medical equipment from mobile phones. J Med Eng Technol 1997;21:141-6.

12. Small DR. Use of mobile phones in hospital. Lancet 2003;361:1831.

13. Klein AA, Djaiani GN. Mobile phones in the hospital – past, present and future. Anaesthesia 2003;58:353-7.

14. Hahn IH, Schnadower D, Dakin RJ, Nelson LS. Cellular phone interference as a cause of acute epinephrine poison-ing. Ann Emerg Med 2005;46:298-9.

15. Calcagnini G, Floris M, Censi F, Cianfanelli P, Scavino G, Bartolini P. Electromagnetic interference with infusion pumps from GSM mobile phones. Health Phys 2006;90:357-60.

16. Lawrentschuk N, Bolton DM. Mobile phone interference with medical equipment and its clinical relevance: a system-atic review. Med J Aust 2004;181:145-9.

17. International Electrotechnical Commission. IEC EN 60601-1-2. Medical electrical equipment. Part 1: General require-ments for safety. 2. Collateral standard: electromagnetic com-patibility. Requirements and tests. Geneva: IEC; 2003.

18. ANSI; Standard C63.18. Recommended practice for an on-site, ad hoc test method for estimating radiated electromagnetic im-munity of medical devices to specific radiofrequency trasmitters. Washington, DC: American National Standard Institute; 1997.

19. International Electrotechnical Commission. IEC EN 60601-2-24. Medical electrical equipment. Part 2-24: Particular re-quirements for the safety of infusion pumps and controllers. Geneva: IEC; 1998.

20. Morrissey JJ. Mobile phones in the hospital: improved mobile communication and mitigation of EMI concerns can lead to an overall benefit to healthcare. Health Phys 2004;87:82-8.